[Nepomuk] Review Request: Ractor ResourceData and ResourceManager

[Commit Hook]

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://git.reviewboard.kde.org/r/105622/#review16313
-----------------------------------------------------------


This review has been submitted with commit 8a1ceb4d3f1fa12e833df0c6a93e7bbacf4d2f54 by Vishesh Handa to branch KDE/4.9.

- Commit Hook


On July 19, 2012, 2:35 p.m., Vishesh Handa wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://git.reviewboard.kde.org/r/105622/
> -----------------------------------------------------------
> 
> Review request for Nepomuk and Sebastian Trueg.
> 
> 
> Description
> -------
> 
> Improve the concepts of kickoff lists and make the code a lot easier to
> understand. Most of the "magic" of the kickoff lists has been removed to
> make it more explicit, and easier to understand.
>     
> Instead of having one kickOff list for everything. We now have 3 lists -
> * One for initialized data. Maps uri -> ResourceData*
> * One for nao:identifiers. Maps nao:identifier -> ResourceData*
> * One for nie:urls. Maps nie:urls -> ResourceData*
>     
> When constructing a ResourceData, the ResourceManager checks the kind
> QUrl it has been given - a nepomukuri, a nie:url or nao:identifier.
> Based on the results it passes the arguemnts to the ResourceData
> constrcutor.
>     
> This change was only possible cause with the DMS, we have made it clear
> that all resource uri's will be of the form 'nepomuk:/'. This makes the
> logic a lot simpler.
> 
> 
> Diffs
> -----
> 
>   libnepomukcore/resource/resourcedata.h 1d0d761 
>   libnepomukcore/resource/resourcedata.cpp 37b48f2 
>   libnepomukcore/resource/resourcemanager.cpp 8f45289 
>   libnepomukcore/resource/resourcemanager_p.h 1547f5e 
> 
> Diff: http://git.reviewboard.kde.org/r/105622/diff/
> 
> 
> Testing
> -------
> 
> I've run the tests, and there are no regressions, and this actually fixes on test.
> 
> But a lot of tests still fail (around 9), so there is still work to be done. This is just the first step.
> 
> 
> Thanks,
> 
> Vishesh Handa
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/nepomuk/attachments/20120724/77389c8d/attachment-0001.html>